Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.

12, 2013

www.iiste.org

Profile Analysis of Students Academic Performance in Ghanaian Polytechnics: The Case of Bolgatanga Polytechnic
Benjamin Apam1* and Albert Luguterah2 Department of Statistics, Bolgatanga Polytechnic, P. O. Box, 767, Ghana, West Africa 2 Department of Statistics, University for Development Studies, P.O. Box, 24, Ghana, West Africa * E-mail of corresponding author: apambenjamin@yahoo.com
1

Abstract The main purpose of the study was to examine the changes in the average academic performance of students over time and how these changes are related to student segments, choice of program and the entry qualification of the student. The cohort of students admitted into Bolgatanga Polytechnic during the 2009/2010 academic year formed the sample and only students who successfully completed were used. Data on grade point averages (GPA), demographic and socio-economic features from 131 Female and 271 Male students was obtained from the Examinations Department and the Student Affairs Unit of Bolgatanga Polytechnic. The multivariate analysis of variance technique was used to complement the Hotellings T2 to compare the mean vectors of k random samples for significant difference among the levels of Departments, Entry Requirements and Gender. Profile analysis of the data indicated at 5% level of significance that the average GPA scores of the Male and Female students were parallel, level and deviated significantly from flatness whereas the various Departments had their own subject-specific mean response. The Entry Qualifications of students admitted into the Polytechnic were not similar. Keywords: Profile Analysis, Academic Performance, MANOVA, Ghana 1. Introduction The Ghanaian Tertiary Education system is composed of the Universities and Polytechnics (including Colleges of Education). The Polytechnic sector emphasizes Technical and Vocational Education by training students in Scientific and Technical subjects. Since 1992, when Government directed Polytechnics in Ghana to run Tertiary programmes, significant gains have been made in the output of the Polytechnic graduate. The Polytechnics provides the bulk of our people with technical education that is relevant, up-to-date in technology, and forward looking in approach (Owusu-Agyeman, 2006). For this reason, Polytechnic graduates play a momentous function in the development of the nation. Many researchers (Abledu, 2012; Gyekye, 2002; Irfan et al.,., 2012; Blazenka et al.,., 2009; Hijazi et al.,., 2006; DeBerard et al.,., 2004; Hansen et al.,., 2003; Deepak et al.,., 2011; Chinwuba et al.,., 2011; Liver et al.,., 2002; Kyei et al., 2011) seem willing to leap from cross-sectional data that describes differences among students academic achievements to making generalisations about change over time. Change is pervasive in everyday life. Beyond the natural changes, targeted interventions can also cause change; test scores or students GPA might rise after coaching, increased learning facilities or students self reported interest in their area of study. By measuring and charting changes like these we uncover the temporal nature of development in terms of students academic performances. This study, however, seeks to describe the changes in the average academic performance of students over time and how these changes are related to student segments, choice of program and entry qualifications. 2. Materials and Methods The study used a cohort of students admitted into Bolgatanga Polytechnic during the 2009/2010 academic year as the study sample. A profile plot on the Grade Point Average (GPA) of students was conducted to establish the growth trajectory and determine the link of its distribution to the exponential family of distributions. 2.1 MANOVA Multivariate analysis of variance evaluates differences among centroids for a set of dependent variables when there are two or more levels of independent variables (groups). This technique provides a multivariate test to compare the mean vectors of k random samples for significant difference when the levels of the grouping variable are more than two. For k independent random samples of size n obtained from pvariate normal populations, the model for each observation is:

yij = + i + ij = i + ij , i = 1 (1) k ;

j = 1 (1) p

(1)

In terms of the p variables in yij, (1) becomes

159

Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.12, 2013

www.iiste.org

y ij1 1 i1 ij1 i1 ij1 M = M + M + M = M + M p yijp ip ijp ip ijp

This model leads to a multivariate hypothesis of the form

11 21 k1 12 22 k2 H 01 : = =L = M M M 1 p 2 p kp
using the likelihood ratio test
H0

( = sup L( , , L ,
1 1 1 H

sup L 1 , 1 , L 1 , x11 , L , x1n1 ,L , x k1 , L , x knk x11 , L , x1n1 ,L , x k1 , L , x knk

) )

where H 0 = (1 , L k , ) : 1 = L = k p , > 0 - - - - - Restricted parameter space H = (1 , L k , ) : 1 = L = k p , > 0 - - - - - Unrestricted parameter space

For the case of two levels of the independent term (group) in the model, the Hotellings T2 was used to test the equality of the mean vectors between the two samples using the following hypotheses: 1. 2. 3.

H 01 : j H 03 : 1

(1)

j
(1)

( 2)

= j 1
( 2) (1)

(1)

j 1 , j = 2, 3,L , p
(2)

H 02 : j
(1)

= j , j = 1, 2,L , p = 1
(2)

= L p

= L p

( 2)

3. Results and Discussion The mean Grade Point Average (GPA) and standard deviations at each measurement location for a random subset of 402 students, broken down by Gender are presented in Table 1. The mean response in the baseline (year one, semester one) was similar across Gender. However, there were discernible differences in the patterns of change in the mean GPA of students over time. As shown in Table 1, the Female students registered the least standard deviation in all semesters except the third semester. According to Table 2, the average GPA for various Departments differed across semesters with the Department of Statistics having the highest mean GPA at the beginning of the first semester and Marketing Department with the least GPA for that semester. In terms of spread across Departments, the Department of Statistics recorded the highest standard deviation for all but the fifth semester whilst the variability of the other departments varied from semester to semester. It was obvious from the descriptive statistics in Table 3 that students who were admitted based on Direct Qualification (DQ) had the highest GPA across all semesters except the third semester. Though, the mean GPA of students who were admitted based on a Decisive factor (NQ) did not change much across semesters, they had a much higher variability as compared to the other two categories. 3.1 Mean Response Profiles of gender Figure 1 indicated a gain in the mean GPA scores for both Male and Female students in the second semester. There was a dramatic drop in the mean GPA scores among Male students in the third semester compared to their Female counterparts. However, each profile had a nonlinear trend as shown in Figure 1. The test of parallelism for the two groups estimated a Hotellings T2 value of 10.4043 with a p-value of 0.0696 which was statistically not significant. Using the MANOVA technique to complement the Hotellings T2, the results indicated at the 5% level of significance that the mean profiles of Male and Female students were parallel 160

Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.12, 2013

www.iiste.org

(Table 4), level (Table 5) and deviated significantly from flatness (Table 6). The evidence of female students performing better than their male counterparts was not seen in this study. This is consistent with the findings of some studies: Chinwuba et al., (2011) established that there was no difference in academic performance between male and female undergraduate Accounting students. Similarly, Ding et al., (2008) in a related study established that ethnicity, but not gender, distinguished two types of change profiles. However, contrary to the results of this study, Deepak et al., (2011) and Hijazi et al., (2006) showed that females performed better than their male counterparts in the field of Medicine and Agricultural Science respectively. The results supporting female dominance in academic performance in subject-based research were done in the developed countries whereas the findings consistent with this study were done in the developing countries. 3.2 Mean Response Profiles of Departments A profile plot of Departments suggested that each Department had their own subject-specific mean response and the observations varied (Figure 2). The results of the multivariate test of parallelism, assuming heterogeneity, indicated at 5% level of significance that the null hypothesis of similar profiles was rejected as shown in Table 7. Hence, the pattern of change of students academic performance differed by departments. The levelness (significance of separation of profiles) and similarities of the response to all the dependent variables, independent of groups were not of much interest once the profiles were not similar. 3.3 Mean Response profiles of Mode of Entry At the beginning of the first academic assessment, students who were admitted with Direct Qualification proved to be much better in terms of academics than those who entered the institution through Matured Entrance Examination (QM) and those that were admitted based on a Decisive Factor (Figure 3). Apart from QD and QM, students who were admitted based on a decisive factor seemed to have a linear trend across all semesters as shown in Figure 3. The matured students, on the other hand, registered the lowest GPA score at all measurement occasions except the second semester. From Table 8, all four multivariate tests rejected the null hypothesis, indicating that the profile of the means were likely different between the three categories of Mode of Entry of the student into the institution. This implies that at least one of the three groups was necessarily not flat. Though it is conceivable that the non-flatness could cancel each other to produce, on average, a flat profile, this result was not of interest to this study. 4. Conclusion The analysis revealed an unsteady linear trend in the mean GPA scores of students in the Bolgatanga Polytechnic. The profiles of the Male and Female students were statistically similar, level and deviated from flatness. However, although the mean GPA score of the Female students was generally higher than their Male counterparts, suggesting a slightly better academic performance in favor of the Female students, this was not statistically significant.. Though the profiles of the various Departments and Mode of Entry were not similar, students admitted on the basis of a decisive factor did not change much across semesters as compared to the non-linear trend showed by students admitted directly and through the Matured Entrance Examination. References Abledu, G. K. (2012). Multiple Regression Analysis of Assessment of Academic Performance of Students in the Ghanaian Polytechnics. Research on Humanities and Social Sciences. Vol. 2, No. 9. Blazenka, D. and Dijana, O. (2009). Prediction of Academic Performance Using Discriminant Analysis. Proceedings of the ITI 2009 31st Int. Conf. on Information Technology Interfaces, Cavtat, Croatia. Chinwuba, A. O., and Osamuyimen, E. (2011). Academic Performance of Male vs Female Accounting Undergraduate Students: Evidence from Nigeria. Higher Education Studies. Vol.1, No. 1. doi: 10.5539/hes.vlnp9. DeBerard, M., Scott, G. I. S. & Deana, C. J. (2004). Predictors of Academic Achievement and Retention Among College Freshmen: A longitudinal Study. College Student Journal 38.1, 66 80. Deepak, K. K., Al-Umran, K. U., Al-Sheikh, M. H., and Al-Rubaish, A. (2011). The Influence of Gender on Undergraduate Performance in Multichoice Testing in Clinical Disciplines at University of Damman, Saudi Arabia. Al Ameen Journal of Medical Science, 4(2), 123-130. Ding, C. S. (2008). Variations in Academic Performance Trajectories During High School Transition: Exploring Change Profiles Via Multidimensional Scaling Growth Profiles Analysis. Educational Research & Evaluation. Vol. 14, Iss. 4. Hansen, T. and Austin, G. (2003). Student Health Risks, Resilience, and Academic Performance in California: Year 2 Report, Longitudinal Analyses. Los Alamitos, CA: WestEd. Irfan, M. and Shabana, N. K. (2012). Factors Affecting Students Academic Performance. Global Journal of Management and Business Research. Vol.12, Issue 9, Version 1.0. Kyei, L., Apam, B. & Nokoe, S. K. (2011). Some Gender Differences in Performance in SHS Mathematics

161

Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.12, 2013

www.iiste.org

Examination in Mixed High Schools. American Journal of Social and Management Sciences. Liver, M. R., Davis-Kean, P. E., & Eccles, J. E. (2002, April). Influences of Gender on Academic Achievement. Presented at the biennial meetings of the Society for Research on Adolescence, New Orleans, LA. Owusu-Agyeman (2006). Government Access-Policies on Polytechnic Education in Ghana: The Relevance for Cape-Coast Polytechnic. Rencher, A. C. (2002). Methods of Multivariate Analysis. 2nd ed. New York: Wiley. Wooldridge, J. M. (2002). Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press. Table 1: Mean GPA (and standard deviation) at each measurement occasion by Gender Gender Female Male Total Semester 1 2.60 (0.59) 2.60 (0.63) 2.60 (0.62) Semester 2 2.67 (0.60) 2.68 (0.62) 2.67 (0.61) Semester 3 2.53 (0.75) 2.36 (0.73) 2.41 (0.74) Semester 4 2.72 (0.61) 2.56 (0.71) 2.61 (0.68) Semester 5 2.60 (0.73) 2.45 (0.80) 2.50 (0.78)

Table 2: Mean GPA (and standard deviation) at each measurement level by Departments DEPARTMENT ACT CVE EAG HCIM IA MKT SMS STA S1* 2.75 (0.52) 2.67 (0.63) 2.33 (0.73) 2.38 (0.58) 2.88 (0.51) 2.28 (0.53) 2.64 (0.71) 2.96 (0.77) S2* 2.85 (0.56) 2.22 (0.50) 2.65 (0.59) 2.35 (0.70) 2.90 (0.45) 2.52 (0.56) 2.61 (0.56) 2.79 (0.75) S3* 2.03 (0.67) 2.39 (0.43) 3.31 (0.45) 2.84 (0.90) 2.86 (0.49) 2.63 (0.57) 2.41 (0.59) 2.73 (0.91) S4* 2.43 (0.70) 2.40 (0.46) 2.85 (0.68) 2.85 (0.69) 3.24 (0.58) 2.65 (0.65) 2.62 (0.51) 2.95 (0.71) S5* 2.28 (0.83) 2.56 (0.64) 2.81 (0.73) 2.36 (0.88) 2.55 (0.62) 2.44 (0.58) 2.95 (0.63) 3.08 (0.84)

* S1, S2, S3, S4 & S5 represent semester1, semester2, semester3, semester4 & semester5

Table 3: Mean GPA (and standard deviation) at semester by Mode of Entry Entry Remark QD NQ QM Semester 1 2.85 (0.64) 2.56 (0.66) 2.53 (0.58) Semester 2 3.00 (0.64) 2.57 (0.67) 2.59 (0.55) Semester 3 2.49 (0.75) 2.57 (0.80) 2.35 (0.72) Semester 4 2.82 (0.66) 2.62 (0.75) 2.54 (0.67) Semester 5 2.68 (0.82) 2.60 (0.79) 2.41 (0.76)

Table 4: Multivariate test of parallelism by Gender Test Statistic Wilks lambda Pillais trace Lawley-Hotelling trace Roys largest root Estimate 0.9775 0.0225 0.0230 0.0230 df 1, 4, 385 4, 385 4, 385 4, 385 F 2.22 2.22 2.22 2.22 p-value 0.0667 0.0667 0.0667 0.0667

162

Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.12, 2013

www.iiste.org

Table 5: Multivariate test of levels by Gender Test Statistic Wilks lambda Pillais trace Lawley-Hotelling trace Roys largest root Estimate 0.9938 0.0062 0.0062 0.0062 df 1, 1, 400 1, 400 1, 400 1, 400 F 2.48 2.48 2.48 2.48 p-value 0.1161 0.1161 0.1161 0.1161

Table 6: Multivariate test of flatness by Gender Test Statistic Wilks lambda Pillais trace Lawley-Hotelling trace Roys largest root Estimate 0.8140 0.1860 0.2285 0.2285 df 1, 4, 398 4, 398 4, 398 4, 398 F 22.74 22.74 22.74 22.74 p-value 0.0000 0.0000 0.0000 0.0000

Table 5: Multivariate test of parallelism by Department Test Statistic Wilks lambda Pillais trace Lawley-Hotelling trace Roys largest root Estimate 0.2928 0.9271 1.7261 1.2649 df 7, 28, 1411.2 28, 1576.0 28, 1558.0 7, 394.0 F 20.46 16.98 24.01 71.20 p-value 0.0000 0.0000 0.0000 0.0000

Table 6: Multivariate test of parallelism by Mode of Entry Test Statistic Wilks lambda Pillais trace Lawley-Hotelling trace Roys largest root Estimate 0.9389 0.0613 0.0650 0.0628 df 2, 8, 788 8, 788 8, 788 8, 788 F 3.16 3.12 3.19 6.20 p-value 0.0016 e 0.0018 a 0.0014 a 0.0001 u

e = exact, a = approximate, u = upper bound on F

163

Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.12, 2013

www.iiste.org

Profile Plot of GPA by Sex

Female Male
2.7

GPA Level

2.3 sem1

2.4

2.5

2.6

sem2

sem3

sem4

sem5

Time (Semesters)
Figure 1: Plot of mean GPA at semester1 to semester 5 in the Male and Female groups

Profile Plot of GPA by Departments

3.5 GPA Level 2 sem1 2.5 3

sem2

sem3

sem4

sem5

Time (Semesters)
ACT EAG IA SMS CVE HCIM MKT STA

Figure 2: Plot of mean GPA at semester 1 to semester 5 by Departments

164

Journal of Education and Practice ISSN 2222-1735 (Paper) ISSN 2222-288X (Online) Vol.4, No.12, 2013

www.iiste.org

Profile Plot of GPA by Mode of Entry

3.2 QD NQ QM mean

Mean GPA Level

2.4 sem1

2.6

2.8

sem2

sem3

sem4

sem5

Time (Semester)
Figure 3: Profile plot of mean GPA levels by Mode of Entry

165

This academic article was published by The International Institute for Science, Technology and Education (IISTE). The IISTE is a pioneer in the Open Access Publishing service based in the U.S. and Europe. The aim of the institute is Accelerating Global Knowledge Sharing. More information about the publisher can be found in the IISTEs homepage: http://www.iiste.org CALL FOR PAPERS The IISTE is currently hosting more than 30 peer-reviewed academic journals and collaborating with academic institutions around the world. Theres no deadline for submission. Prospective authors of IISTE journals can find the submission instruction on the following page: http://www.iiste.org/Journals/ The IISTE editorial team promises to the review and publish all the qualified submissions in a fast manner. All the journals articles are available online to the readers all over the world without financial, legal, or technical barriers other than those inseparable from gaining access to the internet itself. Printed version of the journals is also available upon request of readers and authors. IISTE Knowledge Sharing Partners EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open Archives Harvester, Bielefeld Academic Search Engine, Elektronische Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial Library , NewJour, Google Scholar